Traders spend months refining entry criteria and almost no time examining when they trade. In crypto that omission is expensive, because the market runs continuously and every hour you trade is an hour you chose. The clock is a variable you control completely, and most people have never looked at it.

Sort your last two hundred trades by hour of day. For a lot of traders that single view explains more of the variance than the setup names they’ve been carefully tagging.

The hour shapes your results through you, not the market

There is a real argument about liquidity and volatility across sessions, and it matters at institutional size. At a $2,000 to $10,000 account trading majors, the spread between 3pm and 11pm is not what’s moving your numbers.

What moves them is the state you’re in at each of those hours. Someone who trades after a full workday is doing analysis at the end of nine hours of other analysis. Decision quality drops with fatigue in a way that feels like nothing from the inside - you don’t notice diminished judgement, because the thing doing the noticing is the thing that’s diminished.

This is why the hour is worth measuring rather than reasoning about. You will not feel the difference. It shows up in the record.

There’s a second mechanism working alongside fatigue, and it’s easier to fix. The hours you trade determine which market conditions you ever see. Someone who only trades 9pm to 1am local time is sampling a narrow slice of the day, over and over. If their strategy suits trending conditions and that window is usually quiet, they will spend months concluding the strategy is broken when the mismatch is one of scheduling. Neither the setup nor the trader is the problem, and no amount of refining entries will surface that, because the entries were never the variable.

What the breakdown usually looks like

Take a trader who works until 6pm and trades in the evening. Bucket the trades and a common shape appears:

Nothing about ETH changes at 11pm. What changed is that the trader is four hours further into their evening, has probably already taken a loss or two, and is now trading a market that will still be there tomorrow with a brain that will not.

The late block also tends to hold the revenge trades, because they arrive after losses and losses accumulate through the session. Your worst hour and your most tired hour are usually the same hour, which means the fix is a schedule change rather than a strategy change.

Do the arithmetic before you believe it

Bucketing invites a mistake: slicing until every bucket is tiny, then reading noise as signal.

Suppose 200 trades spread across five two-hour buckets, averaging 40 trades each. A bucket showing a 38% win rate against an account average of 50% looks alarming. On 40 trades, the 95% confidence interval around that 38% runs roughly from 24% to 54%. It overlaps the average. You cannot yet conclude anything.

Two things make the reading trustworthy. Use wider buckets - three or four of them, not eight. And look at more than win rate, because win rate alone hides size. A bucket where you win less often but lose smaller can be fine. A bucket where the win rate holds but average loss doubles is the one quietly draining the account.

The measures worth checking per bucket: win rate, average R-multiple, trades per session, and average holding time. When three of the four move together in the same bucket, you have something real.

Separate the hour from what happened before it

There’s a confound worth handling before you blame the clock. Late trades are not only late - they’re also the trades that come after everything else in the session. If you lost at 9pm, the 11pm trade carries that loss with it.

So the question splits in two. Does the late block underperform on days you were already down, or does it underperform on every day? Those point to different fixes. The first is a loss-response problem and the hour is incidental; capping trades after a losing session would address it. The second is genuinely about fatigue, and only a schedule change helps.

Checking this takes one extra filter. Look at your late-block trades on days where you were flat or up going into that block. If they hold up fine, the clock was never the issue - the sequence was. If they’re still poor on green days, the hour is doing real damage on its own.

Most traders find some of both, which is worth knowing, because it means neither fix alone gets the whole benefit.

Also check the hour you never trade

The reverse question gets asked less often and sometimes pays better. Where are the gaps in your day, and are they gaps by choice or by habit?

Someone trading 7pm to 1am has decided, usually without deciding, that mornings don’t exist. That might be right - work makes it impossible, or the setups they trade don’t appear then. It might also be a default that was never tested against anything.

If your schedule allows even a narrow morning window, twenty trades in it will tell you more than another two hundred in the block you already know. Data you don’t have is worth more than data you have a lot of, and the cheapest experiment available is usually the hour you’ve never measured.

Turn it into one change, not a rule set

The temptation on seeing a bad block is to write elaborate conditions about when trading is permitted. Elaborate conditions get abandoned in week two.

Make one change instead. Take your worst two-hour block and stop trading in it for two weeks. That’s the whole intervention. Two weeks later, compare: total P&L, number of trades, and whether the trades you did take were better than the ones you gave up.

Traders who try this often find the account improves while activity drops by a third, which is uncomfortable if you’ve been treating screen time as effort. It’s also the cheapest edge available, because it costs nothing to implement and requires no new skill.

If cutting the block outright feels too strong, cap it instead: no more than two trades in that window. You’ll learn nearly as much, and you keep the habit of showing up.

Try this tonight

Open Trady and go to the Insight tab. Group your trades by hour and look at the three measures together - win rate, average R-multiple, and count. Ignore any bucket holding fewer than 30 trades; it can’t tell you anything yet.

Find the worst block. Write the two hours on a sticky note and put it on your monitor. Then trade the next fortnight without it, and let the record decide whether that hour was ever earning its place.

Trading carries risk. This article is about analytics, not financial advice.